However, what people often miss is the instrumental argument that economists often make, viz. that income is extremely highly correlated with these "other things" and makes an excellent simple summary statistic.
For a great example of this kind of argument, see Srinivasan, T.N., 1994, Human Development: A New Paradigm or Reinvention of the Wheel? American Economic Review 84(2): 238-243.
Or you could just check out the awesome graph that Justin Wolfers just made:
That's right people, the rank correlation between the UN's Human Development Index and good old income per capita is .95 (out of a maximum of 1.0).
GDP: It's not just for breakfast anymore!
4 comments:
Just out of curiosity, why would Mr. Wolfers choose to put "high" on the inside of both axes, and "low" on the outside of both axes?
Nice graph.
i thought the same thing...the numbers do go 'up', but top right is worst on both areas. Obviously doesn't matter, as the correlation is shown clearly, it's just strange.
I initially thought it went the other way and assumed the small outlier labeled SWZ was Switzerland and was puzzled because I thought Switzerland had a higher income rank than that. I finally looked at it closely and realized that SWZ must be Swaziland.
Forget rank; just doing a linear the UN HDI against the log of GDP per capita gives an adjusted R^2 of 0.88.
Good enough for intergovernmental work?
Post a Comment